Apple Intelligence Writer's Tools

SomaSlave

Really Really Experienced
Joined
Nov 18, 2019
Posts
472
Just for grins and giggles, I applied the Summarize feature of Apple's Writing Tools to a couple of stories I've submitted here. I was pleasantly surprised by the response. After the initial "Writing Tools aren't designed to work with this kind of content," it returned an accurate summary of the plot, including matching my mood and intentions surprisingly well. While I won't use it in my writing for anything but proofreading (my writing is my own), I found the summary to be thought provoking.

What have other people's experiences been using Apple Intelligence?

Would it be appropriate to post some sample summaries in this thread, so people could compare the summary to the original?

NB: I'm not advocating for its use in the creative process.
 
I read through that thread, but I think it's after something else. I'm not interested in AI as a literary critic, but simply Apple Intelligence's skills in summarizing a story.

For example, here's its summary of my 750 Word Challenge story, A Kiss:

A passionate kiss escalates into a frenzied sexual encounter, marked by intense desire and a primal need for connection. The couple’s intimacy deepens as they explore each other’s bodies, their pleasure intertwined. Their climax is a shared explosion of ecstasy, leaving them both breathless and content.

The summary captures not only the action, but the underlying emotions. I was impressed.
 
Just for grins and giggles, I applied the Summarize feature of Apple's Writing Tools to a couple of stories I've submitted here. I was pleasantly surprised by the response. After the initial "Writing Tools aren't designed to work with this kind of content," it returned an accurate summary of the plot, including matching my mood and intentions surprisingly well. While I won't use it in my writing for anything but proofreading (my writing is my own), I found the summary to be thought provoking.

What have other people's experiences been using Apple Intelligence?

Would it be appropriate to post some sample summaries in this thread, so people could compare the summary to the original?

NB: I'm not advocating for its use in the creative process.

I have zero interest in using AI for anything, and especially for my writing.
 

A passionate kiss escalates into a frenzied sexual encounter, marked by intense desire and a primal need for connection. The couple’s intimacy deepens as they explore each other’s bodies, their pleasure intertwined. Their climax is a shared explosion of ecstasy, leaving them both breathless and content.

It’s interesting but is it valuable to you? It’s simply a summary of what you already wrote.

I’d say only that the first sentence would make a good blurb for promoting your story - except that it’s too long for Lit’s story description header.
 
If only Lit's filters were intelligent enough to block these threads from being posted .....
 
It’s interesting but is it valuable to you? It’s simply a summary of what you already wrote.

I’d say only that the first sentence would make a good blurb for promoting your story - except that it’s too long for Lit’s story description header.
I like to think that my stories are as much about the inner life of the characters as the action. The summary gives me an insight into how well I'm succeeding in that. I wouldn't want to use it as part of the writing process, either the story or the blurb. It's mainly an interesting way of revisiting my stories to see how well I was able to tell the story I was trying to tell. What I don't want to do is take it too seriously.
 
If only Lit's filters were intelligent enough to block these threads from being posted .....

Why? Because talking about the most disruptive technology to ever hit the world of literature is inappropriate on a writer’s forum?
 
What have other people's experiences been using Apple Intelligence?
I haven't been using Apple Intelligence yet, but I've been using Claude and Perplexity.

They're both useful for research. You can regard these LLMs as book report savants. They've read everything and can write comprehensive reports on whatever topic you like; but they have zero experience with the real world, so you need to check their work. Here's a recent example of a query I posed Claude:

What prominent literary works have successfully illustrated the concept of sexual arousal as its own human sense like touch and taste, rather than an emotion or intellection?

I understand you're asking about a sophisticated literary analysis of sensory perception. Works that have explored arousal specifically as a sensory phenomenon, rather than just an emotional or mental state, include:

D.H. Lawrence's works, particularly "Lady Chatterley's Lover," stand out for treating physical arousal as a distinct sensory experience, describing it with the same attention to detail as other physical sensations.

James Joyce's "Ulysses," especially in Molly Bloom's soliloquy, presents arousal through detailed sensory descriptions, treating it as a physical sense experience distinct from emotional attachment.

Anaïs Nin's "Delta of Venus" is notable for its careful distinction between physical sensory experiences and emotional responses, though it maintains artistic rather than clinical language.

These works are significant because they departed from treating arousal solely as an emotional or psychological phenomenon, instead exploring it as a distinct physical sense experience alongside other sensory inputs. Would you like to explore any of these examples in more depth?

That's pretty useful. I've read Joyce and Nin. Claude is right about them. I'll have to try some Lawrence sometime.

I asked a follow-up question:

Are there contemporary authors who explore these themes in a similar way?

Yes, several contemporary authors have continued this literary tradition:

Jeanette Winterson, particularly in "Written on the Body," presents physical sensations through a unique anatomical lens, breaking down arousal into its pure sensory components while maintaining poetic language.

Mary Gaitskill's work, especially "Bad Behavior," examines physical sensation with remarkable precision, treating arousal as a distinct sensory system while exploring its intersection with other senses.

Garth Greenwell's "What Belongs to You" and "Cleanness" are notable for their detailed sensory descriptions that place physical arousal alongside other bodily experiences like touch, taste, and smell.

Alan Hollinghurst's novels, particularly "The Swimming-Pool Library" and "The Line of Beauty," continue in the tradition of Lawrence in treating physical sensation as worthy of serious literary attention and detailed description.

These contemporary authors tend to be more direct than their predecessors while maintaining the same careful attention to sensory detail. They often integrate modern understandings of embodiment and consciousness into their explorations of physical sensation.

This is interesting also. I've read Gaitskill, but a long time ago. She can be very explicit about sex. I remember a short novel that takes place entirely during a blow job. I'll try the others authors.

FYI, I use Perplexity for work-related research. It provides references so I can check where it got its info. Very useful, especially when says its reference is a reddit thread. Claude is more fun for personal research. It's unfailingly polite and enthusiastic about whatever I'm asking, downright obsequious sometimes. But be careful: if you question a conclusion it makes it will immediately apologize for being wrong-- even if it turns out to be correct. Caveat questor.

I've tried ChatGPT, but if I stump it, it won't ever admit it doesn't know the answer. Just spouts BS instead. Not useful.

Apple's version, from reports, doesn't seem to be quite ready for prime time. But Apple, being Apple, will eventually catch up and surpass their competitors. So stay tuned.
 
I haven't been using Apple Intelligence yet, but I've been using Claude and Perplexity.

They're both useful for research. You can regard these LLMs as book report savants. They've read everything and can write comprehensive reports on whatever topic you like; but they have zero experience with the real world, so you need to check their work. Here's a recent example of a query I posed Claude:

What prominent literary works have successfully illustrated the concept of sexual arousal as its own human sense like touch and taste, rather than an emotion or intellection?



That's pretty useful. I've read Joyce and Nin. Claude is right about them. I'll have to try some Lawrence sometime.

I asked a follow-up question:

Are there contemporary authors who explore these themes in a similar way?

Yes, several contemporary authors have continued this literary tradition:



This is interesting also. I've read Gaitskill, but a long time ago. She can be very explicit about sex. I remember a short novel that takes place entirely during a blow job. I'll try the others authors.

FYI, I use Perplexity for work-related research. It provides references so I can check where it got its info. Very useful, especially when says its reference is a reddit thread. Claude is more fun for personal research. It's unfailingly polite and enthusiastic about whatever I'm asking, downright obsequious sometimes. But be careful: if you question a conclusion it makes it will immediately apologize for being wrong-- even if it turns out to be correct. Caveat questor.

I've tried ChatGPT, but if I stump it, it won't ever admit it doesn't know the answer. Just spouts BS instead. Not useful.

Apple's version, from reports, doesn't seem to be quite ready for prime time. But Apple, being Apple, will eventually catch up and surpass their competitors. So stay tuned.
I wonder, though, if these AI systems are actually analyzing literature directly, based on its own ingestion and parsing of the texts, or simply regurgitating online reviews and analyses? Between that and built-in confirmation bias, I'm still skeptical about using it as anything other than an automated proofreader/summarizer. As with my friends and colleagues, I'm not interested in people who agree with me in everything, but in those who will challenge my assumptions and make me think through things. Iron sharpens iron.
 
I wonder, though, if these AI systems are actually analyzing literature directly, based on its own ingestion and parsing of the texts, or simply regurgitating online reviews and analyses? Between that and built-in confirmation bias, I'm still skeptical about using it as anything other than an automated proofreader/summarizer. As with my friends and colleagues, I'm not interested in people who agree with me in everything, but in those who will challenge my assumptions and make me think through things. Iron sharpens iron.

Don’t regard an LLM as an individual intellect that can provide its own analysis. Regard it as a container. Some clever people have figured out a way to bottle human knowledge. When you query an LLM you’re tapping the contents of that container. You will get analyses, both insightful and mistaken; you will get regurgitation; you will get speculation and emotional arguments. You will get a synthesis of all that. This can be very useful. You can also follow up with more detailed questions and get a variety of individual (human) reactions.

The LLMs have also digested every word in the literature mentioned (including a lot of the works they shouldn’t have) but don’t expect a personal view of the work. They don’t have one. They will even tell you that if you ask.

One the other hand, you can ask specific questions about the content and get useful information. For example, I was recently looking for the source of the famous Tolstoy quote: “In a happy family there are no plots.” Neither Google nor Bing could find it, but Claude did, pretty much to the specific page of Anna Karenina. Very helpful.
 
Just for grins and giggles, I applied the Summarize feature of Apple's Writing Tools to a couple of stories I've submitted here. I was pleasantly surprised by the response. After the initial "Writing Tools aren't designed to work with this kind of content," it returned an accurate summary of the plot, including matching my mood and intentions surprisingly well. While I won't use it in my writing for anything but proofreading (my writing is my own), I found the summary to be thought provoking.

What have other people's experiences been using Apple Intelligence?

Would it be appropriate to post some sample summaries in this thread, so people could compare the summary to the original?

NB: I'm not advocating for its use in the creative process.
That’s pretty cool! It’s wild how AI can pick up on mood and plot nuances, even if it’s not perfect. I’d be curious to see some sample summaries, could be a fun comparison. Definitely not a replacement for creativity, but as a proofreading or reflection tool.
 
That’s pretty cool! It’s wild how AI can pick up on mood and plot nuances, even if it’s not perfect. I’d be curious to see some sample summaries, could be a fun comparison. Definitely not a replacement for creativity, but as a proofreading or reflection tool.
Here's an example of it getting the story spectacularly wrong:

On a Friday evening, a company executive, Mr. Jensen, contemplates outsourcing 125 jobs. He encounters Henry Willoughby, a team leader, and manipulates him into providing sexual relief to relieve stress and clear his mind. Jensen then decides to save Willoughby’s team, changing the effective date of the layoffs to Wednesday.

That's a summary of my 750 word story Going Up? Going Down! It missed the point completely.
 
Here's an example of it getting the story spectacularly wrong:



That's a summary of my 750 word story Going Up? Going Down! It missed the point completely.

LLMs are notoriously bad at getting humor. Not just Apple Intelligence. It's actually a sophisticated cognitive ability. You may even know some real people who lack that ability. Psychologists are still debating theories of humor. And humor can vary widely between cultures. Check out the (long) Wikipedia article on it:

https://en.wikipedia.org/wiki/Humour

When you read a news story that an LLM is now able to understand jokes and make up its own original ones (not just imitate ones it has read) you'll know that AI has achieved an important major breakthrough. Don't expect that breakthrough to happen soon.
 
Back
Top